How should you measure office space utilization?

Space utilization describes use relative to a defined resource, such as seats or available room time. Density measures office use at room and portfolio levels so workplace teams can decide where to reallocate space. Its workplace utilization examples show how those measurements inform a space plan.

The first step is to state what your percentage measures. A room can be occupied for much of the day while using only a small share of its seats.

What is the difference between occupancy and utilization?

For the calculations below, occupancy is the number of people present at a given moment. Utilization puts use in context by dividing it by a stated capacity or time allowance. Keep those denominators visible whenever you share a chart.

Capacity utilization at a moment

Divide the number of people present by the usable seat capacity. This answers: how full is this space?

Average capacity utilization

Take a time-weighted average of capacity utilization across the stated observation window. This answers: how full is it over time?

Occupied-time rate

Divide time with at least one person present by valid observed available time. This answers: how often is this room used?

Check whether the report measures occupied minutes or qualifying hours

Two percentages can differ even when they use the same observations. One report may divide occupied minutes by available minutes; another may count each clock hour that meets a minimum-use rule.

Consider this hypothetical two-hour window with complete observations. A room is occupied from 09:05 to 09:25 and from 10:05 to 10:35, and empty otherwise. Its occupied-time rate is 50 ÷ 120, or about 41.7%.

Now apply a different, hypothetical rule: count a whole clock hour as used when it contains at least 25 occupied minutes. The first hour does not qualify; the second does. That report shows one qualifying hour out of two, or 50%.

The second result does not mean the room was occupied for 60 minutes. It means one hour met the specified threshold. Keep the measure's name, threshold, clock-hour boundaries and denominator beside the chart. Use the same definition when comparing suppliers or reporting periods, or calculate a common measure from suitable underlying observations.

Peak occupancy

Use the highest observed people count in the stated window. This does not include people who wanted to use the space but could not.

Density's API use-case guide distinguishes utilization relative to capacity from time-used measurement.

Two rooms can look equally busy and need different changes

Consider this hypothetical example. Two ten-seat rooms are available for eight hours. Both have complete observations, with someone present for six hours and nobody present for two.

Room A holds two people throughout those six hours. Room B holds eight.

Both rooms have an occupied-time rate of 6 ÷ 8 = 75%. Across the full eight-hour window, their average capacity utilization differs:

• Room A uses 12 person-hours out of 80 available seat-hours, or 15%.
• Room B uses 48 person-hours out of 80 available seat-hours, or 60%.

Room A is a candidate for investigating smaller meeting spaces. Room B supports a stronger case for keeping group capacity. Examine the pattern across comparable working days and ask employees what they need before changing either room.

How can you tell whether an office has too much space?

Start with the decision area, whether a floor, building or neighborhood. Define its usable capacity, working hours and observation period. Compare typical use with busy periods, then inspect room and desk patterns within that area.

Keep missing observations separate from empty space. If coverage changes, disclose it and use comparable valid periods. Retain the timezone, capacity assumptions and room boundaries so another person can reproduce the calculation.

Look for a practical change you can test. A floor with spare desks may still need its meeting rooms. Check the resources teams depend on before consolidating space, then measure again after the change.